Michalis Avraam » Spatial Data

What is geoprocessing?

Michalis Avraam — Wed, 17 Mar 2010 01:25:23 +0000

An interesting question came about a few days ago. While I have been discussing ESRI’s geoprocessor and how one can use it effectively, I failed to define what geoprocessing is,, either within the ESRI realm or the general GIS realm. In short, geoprocessing is an operation performed on geographic (spatial) data. In other words, it is when one uses data to perform some operations and receiving results. An example would be geocoding. The user provides spatial data (an address), which we geoprocess to identify the output (latitude and longitude). In the ArcGIS world, there are a few methods for geoprocessing: running commands from the ArcToolbox, the Model Builder, the command line and Python scripting (the one I focus on mostly in this blog).

The ArcToolbox

The ArcToolbox of ArcGIS 9.3

The ArcToolbox is a collection of tools provided by ESRI that achieve a number of geoprocessing operations. They are logically organized in themes based on the functionality, and are available to use without any customization. These operations can be very powerful and achieve great results. As you can see from the image on the right, there is a number of tools available to you to use. Depending on the extensions you have installed, you will see some different results, but the Geocoding Tools we mentioned above are available with any kind of license. To use any of these geoprocessing tools, you simply double-click on its name, and a window will pop up asking you for the required inputs and outputs. Pay attention to all the options available to you, and make use of the help button to clarify what the operation does, and how it is achieved.

The Model Builder

The Model Builder Window

The Model Builder is a wonderful program written by ESRI that allows you to graphically define geoprocessing operations. Using a simple tool, a user can drag and drop operations and data on a canvas, and connect them with directional arrows (signifying direction of data). On the example on the right, we follow the Address Geocoding operation. There are three data inputs for the operation, including an address locator, address fields to be used, and finally the input table with the addresses. There is one more data element in the picture, the output feature, on the far right. Note that all data are represented by ovals. The operation, which is in the middle, receives data from all the arrows pointing to it, and returns data to the arrow pointing outwards. Notice that the operation is in a rectangular box rather than an oval, to distinguish it.

The Command Line

The Command Line Window

The command line window is one more way to achieve geoprocessing when one is using ArcGIS. The window simply allows one to type commands, an hit enter to execute them. Some users are more familiar with this interface, and like the speed of which you can type operations. Typing a few letters in the top box will give you a list of all commands starting with those letters. Pressing space will bring the tooltip shown in the top of the image on your right. As you can see, it tells you what the tool expects to run. In this case, it expects an input table (in_table), an address locator (address_locator), the address fields (in_address_fields) and where to save the output (output_feature). Optionally, you can set the out_relation_type (notice the curly braces this time).

Python Scripting

Python Scripting Geocoding

The perhaps most complex way to perform geocoding is done through Python scripting. The example on the right actually presents geoprocessing code. What happens when one uses Python Scripting to create geoprocessing operations is similar with the command line option talked about earlier, with some slight changes. Python scripting gives you the option to interact with the data deeper. What this means is that the user can now access an individual point by latitude and longitude. And since all vector spatial data are built from points, this means you have deep access to any spatial data using Python (to see more about Accessing Geometries using the Geoprocessor click here). Also, as Python is a general use scripting language, you can find reusable code by other users (in the form of modules) to achieve things like reading Excel files (click here for an overview).

And that is what geoprocessing is. Operations that work on spatial data. In the ArcGIS realm, that means either using ArcToolbox, the Model Builder, the Command Line interface or creating your own Python Scripts to achieve your desired results.

Note: There is a rating embedded within this post, please visit this post to rate it.

Understanding the Geoprocessor Programming Model part 1

New Google Maps Lab features

Michalis Avraam — Fri, 12 Feb 2010 22:23:54 +0000

Google Maps introduces some nifty new features for its users in the form of Labs (as in their famous Gmail Labs). The new features are accessible on the toolbar on the top right of the screen, labeled as “New” (look like this: ). There are many features for all map geeks available, including Drag and Zoom (allows you to draw a rectangle on the screen to zoom in it), Aerial Imagery (a much finer resolution imagery than satellite), Rotation (rotates the map, making up South or East), and others. My favorite one is of course the option to introduce the Beta label to Google Maps, for those that suffer from nostalgia.

The image below shows how the labs features are presented:

Google Maps API now supports multiple languages

MapChat 2 Released to Public

Michalis Avraam — Mon, 21 Dec 2009 19:20:13 +0000

MapChat 2

MapChat 2, a wonderful spatial mapping web-based application has been publicly released. This tool enables spatial discussions to enhance the production of local knowledge and public discourse. A demo version is set up for users, so go ahead and give it a spin. There are many features in this release, documented below.

Threaded chat/discussion.
The ability to attach arbitrary spatial footprints to chat messages.
Continual discussion over specific spatial footprints.

The tremendous value of this application comes from its power to enable discussions not only about place, but also in place. The discussions can begin about a specific feature on the map (attached to existing spatial footprint), or about a feature that is not currently present (a new spatial footprint of significance to a user). These discussions can then lead to an agglomeration of spatial footprints of features on the ground that no one has collected data for yet, like play areas for children, or hiking trails, or anything else.

Congratulations to the MapChat team for this accomplishment.

What is geoprocessing?

Google Spatial Data in-house – Goodbye to TeleAtlas?

Michalis Avraam — Thu, 08 Oct 2009 23:43:10 +0000

As mentioned on the Google Geo Blog yesterday, as well as by James Fee in his blog, Google’s spatial offerings (Google Maps and soon Google Earth I assume) will include comprehensive parcel data that has been collected somehow from Google. While the Geo Blog entry is sparse, there is a mention of the “Report a Problem” link added to maps, allowing users to make suggestions (and perhaps corrections) to the data served by Google. What many people have noted is that the copyright notice in the United States has changed, and now the notice reports as a copyright holder Google.

Google Street View Car in Chinatown, Toronto, courtesy of Wikipedia

There is a lot of speculation about the data, including where Google may have procured the data. There are some simple answers for this, as well as many unanswered issues remaining. To tackle it first, there is one great source for road data for Google, and that is their Google Street View Vans/Cars. Those cars are equipped with a good enough GPS to enable the correct georeferencing of all the images taken. Those coordinates, bundled with some OCR scanning Google can do to find street names and the laser range scanners to identify width of road can ultimately assist Google in creating a dataset that may be even more automated than the traditional TeleAtlas methods.

Of course, information like parcels is not easy to achieve using such a method. As James Fee speculated, this can perhaps come from local governments and institutions, but there are multiple problems there as well. Oftentimes these organizations offer their data free online for non-commercial usage. Is Google Maps and Earth commercial if they bring in revenue? Most often. the data require some sort of monetary transfer between a user and the organization. Has Google bought all the data, or received the data for free in some way for exchange of services?

Regardless of the issues, Google seems to have managed what was thought to be improbable: collecting data from a great magnitude of small governments, compiled them and provided them online for people to use free of charge. Whether this will be viewed as evil in the future or not remains to be seen, but for now, Google seems to be doing better than all government initiatives for data interoperability and single warehousing solutions.

Google Maps API now supports multiple languages